Sampling a random hypersphere point from a von Mises-Fisher distribution around the state and retrieving its nearest neighbor asymptotically reproduces Boltzmann exploration probabilities at sublinear cost.
Z ˜V ∈SV orono¨ı(A|Xn+1) fvMF(A | V, κ)κ⟨V, ˜V − A⟩ d ˜V # (86) = fvMF(A | V, κ)κ ⟨V, E Xn∼U (S d−1)
1 Pith paper cite this work. Polarity classification is still indexing.
1
Pith paper citing it
fields
cs.LG 1years
2025 1verdicts
CONDITIONAL 1representative citing papers
citing papers explorer
-
Exploring Large Action Sets with Hyperspherical Embeddings using von Mises-Fisher Sampling
Sampling a random hypersphere point from a von Mises-Fisher distribution around the state and retrieving its nearest neighbor asymptotically reproduces Boltzmann exploration probabilities at sublinear cost.